Preterm delivery, low birth weight and small for gestational age (SGA) are major causes of infant mortality and severe morbidity in the U.S. For African Americans, these risks are nearly doubled. Several studies have implicated air pollution, especially particulate matter <2.5mm (PM2.5) as a risk factor for adverse birth outcomes. A challenge in this research is that particles vary widely in chemical composition by region and season. Due to lack of scientific evidence on which types of particles are most harmful, particles are regulated by size. We will investigate two aspects of the chemical composition of particles: 1) levels of individual PM2.5 chemical components (e.g., nickel); and 2) levels of PM2.5 from specific sources (e.g. oil combustion), to investigate the relationship between exposure to specific types of particles and birth outcomes. We obtained filters from 8 CT and MA monitoring sites that were used to measure PM2.5 total mass. We will analyze these filters to determine levels of 49 chemical components. This chemical component data will be used in source apportionment modeling to estimate PM2.5 from specific sources (e.g. motor vehicles). Our primary outcomes will be birth weight (continuous), low birth weight <2500 gm, very low birth weight <1500 gm, preterm birth <37 wks, very preterm birth <32 wks, and small for gestational age (<10th percentile weight for gestational age), using CT and MA birth certificate data for births (2001-2006) within 30 km of the monitoring sites (N=213,000). Associations will be assessed between birth outcomes and PM2.5 chemical components (Aim 1) and between birth outcomes and the levels of PM2.5 from particular sources (Aim 2). In Aim 3 we will examine traffic PM2.5 with three exposure methods: 1) the chemical component most associated with traffic from Aim 1; 2) traffic PM2.5 levels from Aim 2; and 3) traffic-related air pollution estimated through a GIS traffic model. In our preliminary studies, risk of low birth weight increased 8-13% per interquartile range (IQR) increase in exposure to specific PM2.5 components. We will have >90% power to detect a 10% increase in risk of any outcome (including very low birth weight, prevalence 1.5%) associated with an IQR change exposure to a specific PM2.5 component or PM2.5 source. Associations will be investigated in interaction models to test whether effects differ by race. Findings from this study will identify the components or sources of PM2.5 with the greatest impact on birth outcomes and could lead to interventions targeted at the sources most toxic to infant health.